The research interest of the paper is investigating how various factors impact annoyance and anger propagation in the network. In general, emotion in a human network propagates via process that in psychology is called emotional contagion. While there already are attempts, some very elaborate, to model emotion contagion, to authors’ knowledge none of them look at it as depending on interaction time. To perform simulations, an agent-based model was developed based on real data. The model allows to input personality and interaction frequency as parameters. As a result, some unintuitive results were acquired. First, it was assumed that maximum anger intensity in the network will grow linearly with Neuroticism value, however, the results showed sigmoid character. Secondly, it was also assumed that depending on interaction time, the decrease will be linear but the simulations show very slight decrease or even peak at the beginning. One of the tasks to supplement this research is measuring and confirming the existence of such predicaments in real life.